A post from Shady Characters

Miscellany № 20: On Typewriters

Recently I was lucky enough to take a break from editing the Shady Characters manuscript with an entertaining trip to New York City to visit Brendan Curry, my editor, and Penny Speckter, to whom dedicated readers will require no introduction. (More on that in a week or two, I hope.) In amongst the various punctuation-related news items that popped up since I was away, however, one story in particular caught my eye: the last new typewriter to be manufactured in the UK came off the production line at the Brother factory in Wrexham, Wales, at the end of November this year.1 It is the end of an era, and one that bears a little attention.

The typewriter is one of those disruptive innovations that have arrived periodically to shake up written communication: the unified Carolingian alphabet of the middle ages; Gutenberg’s moveable type of the 1450s; the automated typesetting of fin de siècle Linotype and Monotype machines, and of course the modern Internet, have all, like the typewriter, irrevocably changed the way we write. In the typewriter’s case, however, has that change been for better or for worse? Its demise has inspires mixed feelings in me.*

On the positive side, there’s something reassuringly straightforward about hitting a key and seeing a letter printed right before your eyes, and I appreciate a good, typewriter-inspired monospaced typeface as much as the next coder. For all the nostalgia bound up in the memory of the typewriter, though, it has a darker side. During the century for which it remained the writing implement of choice, those office workers and writers who tapped away ceaselessly at their keyboards were shackled by the typewriters every bit as much as they were emancipated by it.

First, and most apposite to Shady Characters, is the havoc wreaked by the “Great Typewriter Squeeze”, a term coined by writer and blogger J.L. Bell to describe the decimation of punctuation marks caused by the new device.2 Early typists were hamstrung by the paltry selection of symbols available on their typewriters: in addition to an uppercase alphabet and the numbers 2–9, the two-row keyboard on Christopher Latham Sholes’ 1867 prototype bore only ; $ – . , ? and / keys,3 while Sholes’ 1878 QWERTY model added only an apostrophe and a colon.4 Typists could not draw upon the utility of the ampersand, asterisk or octothorpe, nor could they list items @ a unit price or inject emotion with a judicious exclamation mark! (Even a parenthesised aside was out of the question.)

For the most part, we still struggle with the legacy of the typewriter’s unthinking war on punctuation: though marks like the manicule (☞) and pilcrow (¶) can be got at with a combination of keyboard gymnastics and an esoteric knowledge of Unicode code points, for practical purposes we are still confined to the hundred-odd characters made trivially available to us on our laptops, tablets and smartphones.

Second is the typewriter’s social impact. Though it was ostensibly “a tool of female emancipation, offering women a respectable line of work in offices”, the typewriter instead became a visible reminder of an impenetrable glass ceiling. As Samira Ahmed wrote for the Guardian:5

[W]hen the British civil service took over operating telegraph and postal offices in the 1870s, the official in charge, Frank Scudamore, sought out women clerks for their typing speed and dexterity. But crucially, Scudamore said the wages: “which will draw male operators from but an inferior class of the community, will draw female operators from a superior class.” Women would spell and type better, raise the tone of the office, then marry and leave without requiring pensions.

Even when the interrobang (‽) arrived on Remington typewriter keyboards in the late 1960s, at least one newspaper editor still saw the world in Scudamore’s terms. The headline for the Kansas City Kansan’s coverage of this momentous typographic event was “Look Girls, a New Key on Typewriter”.6

None of this will stop me trawling eBay for a pristine IBM Selectric, or dissuade me from wondering if a typewriter is a viable text-input device in the 21st century,7 but it is certainly food for thought. During its heyday the typewriter too was mightier than the sword, and it was double-edged.



Bell, J L. “Dash It All!”.


Weller, C. E. “Home of First Typewriter”. In The Early History of the Typewriter, 20-21.




Ahmed, Samira. “Typing – it’s Complicated”. The Guardian.


Oakley, Don. “Look, Girls, a New Key on Typewriter”. The Kansas City Kansan.


Lowry, Cheryl. “Strikethru”.


“Demise” is perhaps a bit strong; one American company named Swintec will still sell you a new typewriter, and doubtless there are others. 
As an exercise for the reader, right-click this page and hit “view source”, and see the computing world as programmers the world over have done for decades! 

8 comments on “Miscellany № 20: On Typewriters

  1. Comment posted by Michael Hurley on

    I think it may be slightly unfair to the typewriter to lay the entirety of the loss of punctuation and ligatures in typography at it’s feet. Certainly, the typewriter had a lot to do with it, but after all, the original Linotype machine had only 72 channels in it’s matrix magazine (that is, it could only produce 72 discrete characters) and even in its heyday had only 90 channels per magazine*. Easily as much printed matter was produced by Linotype as it was by typewriter. Basically every newspaper and most books were printed from Linotype-cast slugs or stereotypes or film made from those slugs.

    By comparison, the Monotype Composition Caster could produce up to 225 characters from a single matrix case (which was a 15×15 grid), but as the Monotype was a comparatively rarer machine I think the effects of the Linotype and typewriter were more keenly felt.

    *At absolute maximum, you could have 124 characters accessible at one time on a Linotype machine but that was only on specialized machines with a secondary, “auxiliary magazine” mechanism that added another 34 channel side magazine beside the main 90 channel magazine. I think these machines with the auxiliary magazines were only introduced in the late 1930s. They were primarily used to produce things like mathematical and chemical texts that required a large number of unusual characters like Greek letters, fractions, etc. I don’t know of any Linotype or Intertype matrix fonts that had the broad range of punctuation and ligatures hand-cast type had provided before.

    1. Comment posted by Keith Houston on

      Hi Michael,

      Well put! The typewriter was certainly not the only factor in driving out some less common marks of punctuation, as you say, and the Linotype and Monotype too are likely responsible to a degree. It is worth noting, however, that both machines supported many ligatures and a larger range of punctuation off the bat, and also that a dedicated operator of either machine could insert non-standard characters into a page: Linotypes allowed “pi” sorts to be manually inserted into a line, and the individual sorts produced by Monotypes could be replaced with different characters as required after casting. (My understanding is that this is how mathematics were set before the advent of Monotype’s 4-line maths system, for example.)

      Of course, this is not to say that a typist could not pencil in non-standard characters in a similar manner, or that some characters were not already anachronistic when the typewriter arrived.

      The lasting legacy of the typewriter (and to a lesser extent, the teletype) is the keyboard that we’re saddled with today. It isn’t that we can’t throw in a pilcrow if we really want to (much like pencilling in a character on a typewritten page), only that the effort required is usually sufficient to dissuade us.

      I hope you enjoyed the article, and thanks for the very considered comment!

  2. Comment posted by Solo Owl on

    Half a century ago, math was set on Monotype machines by specially trained operators in very, very few shops. At that time there were over 1000 special characters used by mathematicians (probably 2 or 3 times as many today). Displayed equations still had to be set by hand, at great expense, with their subscripted superscripts &c and special rules for kerning and spacing &c. No doubt this is why equations and even %-signs were (and are) absent from newspapers and most books.

    (The introduction of computer typesetting actually made things worse, at first. It required the genius of Donald Knuth to create a computer program that could replace quality hand composition. His program, modified a bit, is still used by nearly all mathematicians, physicists, chemists, and computer scientists, and many linguists.)

    1. Comment posted by Keith Houston on

      Hi Solo Owl,

      It’s interesting you should mention this — I’m working on a chapter about Linotype and Monotype machines for The Book as we speak, and one of the points I’m exploring is the MT’s greater flexibility in setting mathematics.

      (La)TeX certainly was and is a huge achievement. I used it in university, though non-academic publishing seems to have plumped for Word as its standard file format.

      Thanks for the comment!

  3. Comment posted by Solo Owl on

    Keith, when you do buy your IBM Selectric, you should start collecting specialized replacement typing balls. I wonder how much of Unicode was anticipated on those little balls.

    1. Comment posted by Keith Houston on

      This is an interesting point! Did type balls have different character sets, or did they all match the keyboard, I wonder?

    2. Comment posted by Solo Owl on

      Some matched the keyboard, and some had entirely different characters. The point was not only to get italic and bold shapes of the same characters, but to greatly expand the number of characters. A technical typist might have 6 or 8 typing balls, so that they would not have to leave large blank spaces for the Greek letters and special signs to be added by hand (not quite a throwback to medieval illumination).

Leave a comment

Required fields are marked *. Your email address will not be published. If you prefer to contact me privately, please see the Contact page.

Leave a blank line for a new paragraph. You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>. Learn how your com­ment data is pro­cessed.